LIDERAZGO ORIENTACIÓN
3.6. Procedimiento en la aplicación de instrumentos
The purpose of the study was to identify characteristics of a conjoint therapy program for IPV with Aboriginal couples. The focus was on service providers’
perspectives of the content and activities for such a program. This thesis utilized existing data previously collected as part of a large-scale study of family violence professionals in 2012. The approved ethics protocol for the complete study is included in the appendix. The author performed the interpretation of multidimensional scaling and cluster analysis results in the concept maps.
As part of the larger study, telephone interviews of experts who provided safety and therapeutic services to Aboriginal couples experiencing IPV were conducted. Participants answered several questions including the two questions which are the focus of this thesis: 1) “How would you address content with Aboriginal men in couple counselling who use abusive behaviour toward their intimate partner?” and 2) “How would you address activities with Aboriginal men in couple counselling who use abusive behaviour toward their intimate partner?”. Results were analyzed using Concept Mapping (Trochim, 1989). In this chapter a brief overview of Concept Mapping and the procedures followed are provided.
Concept Mapping
Concept Mapping is a method used to analyze and interpret qualitative data (Chambers, 1992). Developed by Bill Trochim (1989), Concept Mapping is a structured conceptualization process, which provides a visual representation of individuals’ views on a particular research topic and their conceptual relationships to one another. This method applies a uniform structure to the analysis of qualitative data. Concept Mapping
has been used for program planning and evaluation (Trochim, Stillman, Clark & Schmitt, 2003), to develop conceptual frameworks (Cousineau, Houle, Bromberg, Fernandez & Kling, 2008) as well as theories of change (Petrucci & Quinlan, 2007).
Procedure
The Concept Mapping process includes six steps: 1) Preparation, 2) Generation, 3) Structuring, 4) Representation, 5) Interpretation and 6) Utilization (Trochim, 1989). The steps for concept mapping were followed in the present study, beginning with the collection of responses to a focal question. Next, participants’ responses were edited and redundant responses were excluded. Individual responses were printed on cards.
Participants were asked to sort the responses into groups. Two statistical procedures were used to analyze the sort data. Multidimensional scaling was used to represent the
responses as points on a point map while cluster analysis combined responses to form concepts. The Concept System (Trochim, 1987) was utilized to perform the analyses. The writer and her advisor determined an appropriate number of concepts and assigned a label to each for the two questions described in this thesis.
Preparation. Individuals were invited to participate in the study if they had expertise in either of the following areas: 1) service provider in safety area (e.g. shelter, law enforcement), or 2) professional providing treatment for IPV (e.g. group facilitator, support services worker). Participants were recruited through advertisement to staff working within justice, family services and health departments in a Canadian province. Interested participants were invited to contact the researchers to arrange for individual telephone interviews.
During the telephone interview participants’ responses to demographic questions as well as to several open-ended questions were noted by researchers. No audio recording was done. The present study focused on two of the open-ended questions: 1) “How would you address content with Aboriginal men in couple counselling who use abusive behavior with their intimate partner?” and 2) “How would you address activities with Aboriginal men in couple counselling who use abusive behavior with their intimate partner?” These research questions were part of the larger study which had received ethics approval from the “Sub Research Ethics Board” within the Education Department at Western
University.
Generation. At the time of interview each was informed that Western
University’s institutional ethics board had approved the study and that participation was voluntary, confidential and each had the right to withdraw at any time without
consequence. All individuals who were contacted agreed to participate. At the end of the interview each was asked if she or he was interested in participating in the sorting task at a later time. A list of interested individuals was kept.
Participants included 26 individuals who provided services, including both safety and therapeutic, for IPV to Aboriginal couples. They were between 29 to 64 years of age (average age = 49). The sample consisted of mainly female participants, with male participants constituting ¼ of the sample. Participants had from 3 to 40 (average = 22) years of experience working in the area of IPV. At the time of interview participants were employed as front line law enforcement and corrections as well as women’s shelter staff, senior administrators in family violence programming as well as experienced therapists providing treatment.
Structuring. All responses provided to each question by participants were separately examined. Unique responses to each question were identified by having three researchers flag responses that were redundant. Any response identified as redundant by at least two researchers was excluded and any responses identified as ambiguous by at least one, was edited. This process resulted in 74 unique responses to question one and 44 unique responses to question two.
Participants who agreed to participate in the sorting task were contacted by telephone to confirm their interest in the sorting task and mailed out packages containing copies of all responses by question, printed on separate cards. Consistent with the
instructions provided by Trochim (p. 5, 1989) on how to complete the sorting task, participants were given the following instructions: a) “a response card cannot be placed in two piles simultaneously”; b) “all responses cannot be placed in a single pile”, and c) “all responses cannot be put into their own piles (although some items may be sorted by themselves)”. Each participant was asked to sort the cards together in a way that she or he found most meaningful (Trochim, 1989).
Representation. The sort data was analyzed using multidimensional scaling and cluster analysis with the Concept System (Trochim, 1987). Multidimensional scaling organized responses as points on a map. Responses that were sorted together more frequently by participants were closer to one another on the map. Responses that were rarely sorted together were further apart on the map. Multidimensional scaling results were used for the cluster analysis of responses which grouped responses into concepts.
Multidimensional scaling. A similarity matrix was constructed based on the sorting results. The matrix indicated the frequency with which any one response was
sorted with every other response (Kane & Trochim, 2007). A similarity matrix was calculated for each individual participant and a combined matrix value was calculated on across all participants (Kane & Trochim, 2007). Using a multidimensional scaling of the similarity matrix on a two-dimensional plot (X-Y graph), responses were placed as points on a map. The point map visually represented the frequency with which each response was sorted with the other responses (Kane & Trochim, 2007).
Cluster analysis. Cluster analysis was used to identify the conceptual structure underlying the responses on the map (Everett, 1980). Ward's technique was used to determine the conceptual clusters on the map. The technique began with each response as its own concept and at each stage two concepts were combined until all responses were in one concept. Ward's method provided solutions that were more meaningful and
interpretable than other cluster analysis techniques (Trochim, 2015).
The bridging index. A bridging index was calculated for each response. The bridging index is a statistical value ranging from 0 to 1. A bridging index value is used to describe the relationship between each response with the other responses near it on the map. A low bridging index (between 0-.25) indicated that the response was sorted with other responses in its close proximity. A high bridging index (between 0.75-1) indicated that the response was frequently sorted with responses in areas farther from it on the map. An average bridging index value was also calculated for each concept. Smaller average bridging index values were indicative of highly cohesive concepts.
Interpretation. Researcher judgment was used to determine the most appropriate number of concepts. In general, maps with fewer concepts provided a broad visual representation of the most important ideas, while maps with greater numbers of concepts
provided a more detailed view (Kane & Trochim, 2007). In selecting the most
appropriate number of concepts as few concepts as possible should be chosen (Trochim, 1989). The writer and her advisor reviewed maps for each question and made the
decision about the most appropriate number of concepts to be included in each map. These decisions were made based on conceptual similarity between responses in the same concept as well as conceptual differences between responses in different concepts. Each concept was then assigned a label to reflect its contents (Trochim, 1989).
Utilization. Concept mapping provided a visual representation of the participants’ main ideas as well as their conceptual structure (Trochim, 1989). In Chapter Four, the maps are presented along with a description of the responses in each map. In Chapter Five the results of the concept maps are compared and contrasted with the current literature.
Chapter 4: Results
The purpose of the study was to identify characteristics of a conjoint therapy program for IPV with Aboriginal couples. The focus was on service providers’ perspectives of the content and activities for such a program. The responses were
collected through telephone interviews with professionals who provided either a safety or therapeutic service for IPV to Aboriginal couples. Responses to two questions were the focus of the present study: “How would you address content with Aboriginal men in couple counselling who use abusive behavior with their intimate partner?” and “How would you address activities with Aboriginal men in couple counselling who use abusive behavior with their intimate partner?” Responses to these questions were then sorted into groups by participants. The sort data was analyzed using multidimensional scaling and cluster analysis. A concept map was created for each question. This chapter presents the results of the concept maps for each question.
Content
The question, “How would you address content with Aboriginal men in couple counselling who use abusive behavior with their intimate partner?” elicited 74 unique responses. Six participants sorted the responses. Multidimensional scaling as well as cluster analysis were used to analyze the sort data. According to Trochim (1989), for 100 unique responses or less, concept maps with 3 to 20 concepts should be examined. This ensures that the final number of concepts in a map accurately capture the most
meaningful themes underlying the data.
For the first question, concept maps reflecting a range of possible solutions were reviewed. Maps containing 20 and 15 concepts showed a significant amount of scatter.
Concept maps with 10 to 4 concepts were then reviewed and assessed for the best data interpretability. The concept map containing ten concepts did not produce clusters that were significantly differentiated. As the number of concepts decreased from nine to five, the map showed greater differentiation between concepts. The concept map containing four concepts did not capture the complexity of the data. The concept map which most accurately represented the data contained six concepts (see Figure 1).
Individual responses along with their corresponding bridging index values are shown in Table 1. The bridging indices helped determine the most important responses belonging to each concept. They were also used to help label concepts. Each bridging index took a value between 0 and 1. A value of 0.0 to 0.25 was considered to be in the low range. Any response with a low bridging value implied that it was frequently
grouped only with responses in the same concept. A bridging index value of 0.75 to 1.00 was considered to be in the high range. A response with a high bridging index value implied that the response was also frequently grouped with responses in other concepts. Trochim (1989) suggests that responses with high bridging index values do not
conceptually fit solely with responses in their immediate surroundings. As such, these responses may be less representative of the general theme of the concept they belong to.
An average bridging value was calculated for each concept. Concepts containing responses that were more consistent with one another had a lower average value. Higher average bridging values indicated that responses in the concept were sorted with
Figure 1. Concept Map for Question One. Table 1
Concept Items and Bridging Values for Concept Map for Question One
Concept and Response Bridging Index
Cultural 0.32
1. Aboriginal healing approaches 0.00
2. Aboriginal spirituality 0.00
9. ceremonies 0.00
15. cultural practices 0.13
68. traditional approaches 0.13
69. traditional values 0.24
14. cultural context 0.30
22. Elders involved 0.32
32. historically this is what the family looked like 0.36
29. healing circles 0.53
18. different beliefs and tribes 0.59
42. language specific 0.61
60. skills based in the process of a circle 0.66
54. ritual 0.70
Western 0.67
4. addictions 0.52
47. pride 0.57
62. strength 0.63
11. Cognitive behavioral therapy 0.63
58. self-esteem 0.64
49. psychological/psychiatric concepts 0.69
20. drugs and alcohol 0.70
6. belonging 0.71
43. mental health 0.74
59. sexism 0.76
Traditional Ways 0.80
13. couples needs 0.55
65. talking and sharing 0.70
70. trauma 0.74
52. residential schools 0.76
39. intergenerational trauma 0.84
28. ground therapy in holistic way 0.88
10. children 0.97
8. broad focus 1.00
What a Healthy Relationship Is (and Is Not) 0.33
67. time-outs 0.22
30. healthy relationships 0.26
16. cycle of violence 0.26
41. knowing what their risk factors are 0.27
46. power and control 0.28
17. dealing with conflict 0.29
21. dynamics of relationship 0.29
63. taking responsibility 0.29
44. ownership of behavior 0.29
33. how his abuse impacts the relationship 0.31
3. accountability 0.32
5. anger 0.34
50. public ownership by him of his abusive behaviour towards her 0.34
12. communication skills 0.35
35. impact of abuse on themselves, partner, and children 0.35
64. talk about forms of abuse 0.37
25. express needs in healthy ways 0.37
31. help them identify abusive behavior 0.38
19. discipline with children 0.39
24. equality 0.39
57. safety 0.49
56. role of male 0.50
Men’s Self Responsibility 0.44
40. jealousy 0.36
71. trust 0.36
72. understand triggers 0.37
23. emotion management 0.43
51. regulate anger 0.47
27. feelings and expectations 0.49
73. violence 0.56
Accountability to Family and Community 0.58
61. stop violence from happening with their children 0.40
34. how to live as a family 0.42
74. what does a healthy community and household look like 0.42
53. respect for women 0.48
45. parenting 0.55
37. incorporate local resources in the community 0.66
55. role modeling 0.71
26. extended families and closeness of families 0.73
36. importance of family 0.82
Cultural. This concept emphasized the importance of integrating elements of Aboriginal culture and practices into IPV treatment. Responses in this concept included: “Aboriginal healing approaches”, “Aboriginal spirituality” as well as getting “Elders involved”, “ceremonies”, “cultural practices”, “skills based in the process of a circle”, “ritual” and “healing circles”. Other responses in this concept included: “traditional approaches”, “blend of western and traditional concepts”, “traditional values” and “cultural context”. Responses such as “different beliefs and tribes” and “language specific” indicated that Aboriginal communities are culturally diverse and that IPV treatment for Aboriginal couples needs to be sensitive to such diversity. The response
“historically this is what the family looked like” suggested that teaching clients about what constituted a normal structure of family within Aboriginal culture (e.g. extended family) should also be addressed with Aboriginal couples.
Western. In this concept, participants provided a number of responses that reflected a Western-based approach to IPV treatment. From a Western perspective “addictions”, “drugs and alcohol” are often predictive of violence and need to be
addressed in IPV treatment. Responses such a “psychological/psychiatric concepts” and “mental health” indicated that from a Western perspective, an effective IPV treatment needs to address the couple’s mental health problems as well. Other responses including “pride”, “strength”, “self-esteem” and “belonging” referred to traits and tendencies that when exaggerated (e.g. narcissistic tendencies) from a Western perspective, they can turn pathological and potentially lead to violence. “Intergenerational cycle of abuse” and “sexism” referred to other risk factors for female partner abuse. Lastly, professionals indicated “cognitive behavioral therapy” which is a Western-based therapeutic approach, can be used to address IPV with Aboriginal men.
Traditional Ways. This concept emphasized the importance of taking a traditional approach to address violence with Aboriginal men. The responses such as “ground therapy in holistic way” and “broad focus” suggested that the treatment of IPV with Aboriginal couples needs to be contextualized and conducted in a holistic manner consistent with Aboriginal traditional ways of healing. The response “couples’ needs” suggested that the treatment sessions need to be structured and guided based on the couple’s needs. The response “talking and sharing”, highlighted the importance of silence and the manner with which the couple and the therapist communicate with one another,
with respect and without interruption. The response “children” indicated that not only the couple, but also the children need to be involved and focused on in the therapeutic work. Other responses in this concept included “trauma”, “residential schools”, and
“intergenerational trauma” which are all factors that uniquely contribute to violence among Aboriginal clients and need to be considered in therapeutic work with these clients.
What a Healthy Relationship Is (and Is Not). This concept highlighted the importance of teaching couples about the characteristics of a healthy and safe relationship as well as the ways of building such a relationship. Participants in this study suggested that violence can be prevented by teaching couples about various concepts such as the “cycle of violence”, “anger”, “dynamics of relationship”, “dealing with conflict”, “power and control”, “equality”, “role of male”. and “knowing what their risk factors are” for violence. To maintain a healthy relationship couples also need to learn about what constitutes “healthy relationships” and learn about specific strategies including “time- outs”, “communication skills”, “problem solving” and learning to “express needs in healthy ways”. Other responses including “ownership of behaviour”, “accountability”, “taking responsibility”, “public ownership by him of his abusive behaviour towards her” highlighted the importance of having men take responsibility for the violence as well as the changes they want to make congruent with a healthy relationship. Responses such as “talk about forms of abuse”, “help them identify abusive behaviour”, “how his abuse impacts the relationship” and “impact of abuse on themselves, partner, and children” suggested that learning about the negative consequences of abusive behaviour can lead to the motivation to end abusive behaviours, thereby facilitating a healthy relationship.
Other signs of a healthy relationship included how safe children and female partners felt in the relationship. Responses such as “discipline with children” and “safety” represented appropriate safe behaviours toward children and female partners that were indicative of healthy relationships among family members.
Men’s Self-Responsibility. This concept highlighted the importance of men’s personal responsibility for violent behaviour, as well as feelings and thoughts that can potentially lead to violence. Responses such as “understand triggers”, “emotion
management” and “regulate anger” implied that men needed to take ownership for their abusive behaviours and the positive change they want to create. Other responses
including “feelings and expectations”, “thoughts and behaviors”, “violence”, “jealousy” and “trust” indicated that men need to take charge of their individual thoughts and feelings that can potentially trigger violence.
Accountability to Family and Community. This concept implied that violence can be addressed with Aboriginal men by increasing the sense of accountability they feel